Summary of Individuals who Completed Post-Injury IMPACT Testing

Table 1

Plot 1

Table 2

Plot 2

Is there a significant difference in IMPACT test performance for individuals who completed baseline and post-injury assessments?

Verbal Memory

One Baseline and One Post-Injury Test

A paired T-test was used to compare the mean verbal memory composite scores of individuals who completed one baseline and one post-injury assessment (n = 4,540). Performance on the verbal memory composite between baseline 1 and post-injury 1 assessments was determined to not be significant, t(4,539) = 1.837, p = .066. The mean baseline 1 score (M = 82.56) was not significantly different than the mean post-injury 1 score (M = 82.19).

Boxplot

T-Test Results

## 
##  Paired t-test
## 
## data:  bl_pi_mem_verbal_1 and bl_pi_mem_verbal_2
## t = 1.8369, df = 4539, p-value = 0.06629
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.02518812  0.77408680
## sample estimates:
## mean of the differences 
##               0.3744493

Baseline Summary

Baseline Histogram

Post-Injury Summary

Post-Injury Histogram

Two Baseline and Two Post-Injury Tests

For individuals who completed two baseline and two post-injury assessments (n = 2,498), performance on the verbal memory composite score was statistically significantly different, F(2.82, 7,030.62) = 244.45, p < .0001, generalized eta squared = 0.05.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. All pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest verbal memory score on the second baseline administration (M = 85.57) and their lowest score on the first post-injury administration (M = 78.60).

Boxplot

Outliers

Of the 139 scores identified to be outliers, 5 are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn     DFd       F         p p<.05   ges
## 1   test 2.82 7030.62 244.451 3.04e-142     * 0.046

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Three Baseline and Three Post-Injury Tests

For individuals who completed three baseline and three post-injury assessments (n = 254), performance on the verbal memory composite score was statistically significantly different across tests, F(4.2, 1,061.98) = 54.55, p < .0001, generalized eta squared = 0.11.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Twelve out of 15 pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest verbal memory score on the third baseline administration (M = 87.50) and their lowest score on the first post-injury administration (M = 75.24). The following pairwise comparisons were not significant: Baseline 1 and Post-Injury 3, Baseline 2 and Baseline 3, and Baseline 2 and Post-Injury 3.

Boxplot

Outliers

Of the 24 scores identified to be outliers, 1 is considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect DFn     DFd      F        p p<.05   ges
## 1   test 4.2 1061.98 54.663 1.11e-43     * 0.108

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Baseline 3 Summary

Baseline 3 Histogram

Post-Injury 3 Summary

Post-injury 3 Histogram

Visual Memory

One Baseline and One Post-Injury Test

A paired T-test was used to compare the mean visual memory composite scores of individuals who completed one baseline and one post-injury assessment (n = 4,540). Performance on the visual memory composite between baseline 1 and post-injury 1 assessments was determined to not be significant, t(4,539) = -0.497, p = .62. The mean baseline 1 score (M = 72.16) was not significantly different than the mean post-injury 1 score (M = 72.27).

Boxplot

T-Test Results

## 
##  Paired t-test
## 
## data:  bl_pi_mem_visual_1 and bl_pi_mem_visual_2
## t = -0.4971, df = 4539, p-value = 0.6191
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.5531866  0.3293980
## sample estimates:
## mean of the differences 
##              -0.1118943

Baseline Summary

Baseline Histogram

Post-Injury Summary

Post-Injury Histogram

Two Baseline and Two Post-Injury Tests

For individuals who completed two baseline and two post-injury assessments (n = 2,498), performance on the visual memory composite score was statistically significantly different, F(2.93, 7,309.61) = 226.65, p < .0001, generalized eta squared = 0.04.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Five out of six pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest visual memory score on the second baseline administration (M = 76.53) and their lowest score on the first post-injury administration (M = 68.80). Average performance on Baseline 1 (M = 72.11) and Post-Injury 2 (M = 72.31) was not significantly different.

Boxplot

Outliers

Of the 72 scores identified to be outliers, none are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn     DFd       F      p p<.05   ges
## 1   test 2.93 7309.61 226.645 2e-137     * 0.039

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Three Baseline and Three Post-Injury Tests

For individuals who completed three baseline and three post-injury assessments (n = 254), performance on the visual memory composite score was statistically significantly different across tests, F(4.7, 1,189.47) = 51.78, p < .0001, generalized eta squared = 0.09.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Twelve out of 15 pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest visual memory score on the third baseline administration (M = 78.76) and their lowest score on the first post-injury administration (M = 66.08). The following pairwise comparisons were not significant: Baseline 1 and Post-Injury 3, Baseline 2 and Baseline 3, and Post-Injury 1 and Post-Injury 2.

Boxplot

Outliers

Of the 12 scores identified to be outliers, none are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect DFn     DFd      F        p p<.05   ges
## 1   test 4.7 1189.47 51.776 3.49e-46     * 0.093

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Baseline 3 Summary

Baseline 3 Histogram

Post-Injury 3 Summary

Post-injury 3 Histogram

Impulse Control

One Baseline and One Post-Injury Test

A paired T-test was used to compare the mean impulse control composite scores of individuals who completed one baseline and one post-injury assessment (n = 4,540). Performance on the impulse control composite between baseline 1 and post-injury 1 assessments was determined to not be significant, t(4,539) = -0.850, p = .395. The mean baseline 1 score (M = 7.73) was not significantly different than the mean post-injury 1 score (M = 7.82).

Boxplot

T-Test Results

## 
##  Paired t-test
## 
## data:  bl_pi_impulse_control_1 and bl_pi_impulse_control_2
## t = -0.8502, df = 4539, p-value = 0.3953
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.3087457  0.1219616
## sample estimates:
## mean of the differences 
##             -0.09339207

Baseline Summary

Baseline Histogram

Post-Injury Summary

Post-Injury Histogram

Two Baseline and Two Post-Injury Tests

For individuals who completed two baseline and two post-injury assessments (n = 2,498), performance on the impulse control composite score was statistically significantly different, F(2.66, 6,633.91) = 17, p < .0001, generalized eta squared = 0.003.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Three out of six pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest impulse control score on the first post-injury administration (M = 8.50) and their lowest score on the second post-injury administration (M = 7.53). The following pairwise comparisons were not not significant: Baseline 1 and Baseline 2, Baseline 1 and Post-Injury 2, and Baseline 2 and Post-Injury 2. Performance on Post-Injury 1 was signficiantly different from all other comparisons.

Boxplot

Outliers

Of the 446 scores identified to be outliers, 78 are considered extreme.

Normality Assumption

From the plot, normality is assumed, although extreme outliers on the Post-Injury 1 assessment are concerning.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn     DFd      F        p p<.05   ges
## 1   test 2.66 6633.91 17.003 5.28e-10     * 0.003

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Three Baseline and Three Post-Injury Tests

For individuals who completed three baseline and three post-injury assessments (n = 254), performance on the impulse control composite score was statistically significantly different across tests, F(3.73, 943.1) = 2.99, p = .021, generalized eta squared = 0.007.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Only 2 out of 15 pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest impulse control score on the first post-injury administration (M = 9.17) and their lowest score on the first baseline administration (M = 7.5). Only the differences in scores between Baseline 1 and Post-Injury 1 and Baseline 2 and Post-Injury 1 were significant.

Boxplot

Outliers

Of the 67 scores identified to be outliers, 17 are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn   DFd     F     p p<.05   ges
## 1   test 3.73 943.1 2.988 0.021     * 0.007

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Baseline 3 Summary

Baseline 3 Histogram

Post-Injury 3 Summary

Post-injury 3 Histogram

Reaction Time

One Baseline and One Post-Injury Test

A paired T-test was used to compare the mean reaction time composite scores of individuals who completed one baseline and one post-injury assessment (n = 4,540). Performance on the reaction time composite between baseline 1 and post-injury 1 assessments was determined to be significantly different with individuals achieving a better reaction time score on the baseline assessment, t(4,539) = -3.893, p = .0001. The mean baseline 1 score (M = 0.64) was significantly different than the mean post-injury 1 score (M = 0.65).

Boxplot

T-Test Results

## 
##  Paired t-test
## 
## data:  bl_pi_reaction_time_1 and bl_pi_reaction_time_2
## t = -3.8927, df = 4539, p-value = 0.0001006
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.011976096 -0.003953419
## sample estimates:
## mean of the differences 
##            -0.007964758

Baseline Summary

Baseline Histogram

Post-Injury Summary

Post-Injury Histogram

Two Baseline and Two Post-Injury Tests

For individuals who completed two baseline and two post-injury assessments (n = 2,498), performance on the reaction time composite score was statistically significantly different, F(2.55, 6,365.51) = 195.75, p < .0001, generalized eta squared = 0.04.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Six out of six pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their best reaction time score on the second baseline administration (M = 0.61) and their lowest score on the first post-injury administration (M = 0.67).

Boxplot

Outliers

Of the 343 scores identified to be outliers, 84 are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn     DFd       F         p p<.05   ges
## 1   test 2.55 6365.51 195.748 2.36e-104     * 0.038

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Three Baseline and Three Post-Injury Tests

For individuals who completed three baseline and three post-injury assessments (n = 254), performance on the reaction time composite score was statistically significantly different across tests, F(3.6, 911.29) = 41.72, p < .0001, generalized eta squared = 0.09.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Twelve out of 15 pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their best reaction time score on the third baseline administration (M = 0.60) and their lowest score on the first post-injury administration (M = 0.70). The following comparisons were not significantly different: Baseline 3 and Post-Injury 3, Baseline 2 and Post-Injury 3, and Baseline 1 and Post-Injury 2.

Boxplot

Outliers

Of the 36 scores identified to be outliers, 7 are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect DFn    DFd      F        p p<.05   ges
## 1   test 3.6 911.29 41.715 1.93e-29     * 0.086

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Baseline 3 Summary

Baseline 3 Histogram

Post-Injury 3 Summary

Post-injury 3 Histogram

Visual Motor

One Baseline and One Post-Injury Test

A paired T-test was used to compare the mean visual motor composite scores of individuals who completed one baseline and one post-injury assessment (n = 4,540). Performance on the visual motor composite between baseline 1 and post-injury 1 assessments was determined to be significantly different with individuals achieving a better visual motor score on the post-injury assessment, t(4,539) = -19.244, p < .0001. The mean baseline 1 score (M = 34.11) was significantly different than the mean post-injury 1 score (M = 36.04).

Boxplot

T-Test Results

## 
##  Paired t-test
## 
## data:  bl_pi_visual_motor_1 and bl_pi_visual_motor_2
## t = -19.244, df = 4539, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.125953 -1.732840
## sample estimates:
## mean of the differences 
##               -1.929396

Baseline Summary

Baseline Histogram

Post-Injury Summary

Post-Injury Histogram

Two Baseline and Two Post-Injury Tests

For individuals who completed two baseline and two post-injury assessments (n = 2,498), performance on the visual motor composite score was statistically significantly different, F(2.61, 6,514.32) = 522.6, p < .0001, generalized eta squared = 0.06.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Six out of six pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest visual motor score on the second baseline administration (M = 37.98) and their lowest score on the first baseline administration (M = 33.92).

Boxplot

Outliers

Of the 119 scores identified to be outliers, none are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn     DFd       F         p p<.05  ges
## 1   test 2.61 6514.32 522.598 1.17e-268     * 0.06

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Three Baseline and Three Post-Injury Tests

For individuals who completed three baseline and three post-injury assessments (n = 254), performance on the visual motor composite score was statistically significantly different across tests, F(3.82, 967.55) = 82.59, p < .0001, generalized eta squared = 0.1.

All possible pairwise comparisons were evaluated using the Bonferroni procedure to control family-wise Type I error. Thirteen out of 15 pairwise comparisons were statistically significantly different (p < .05). On average, individuals achieved their highest visual motor score on the third baseline administration (M = 39.42) and their lowest score on the first post-injury administration (M = 32.67). The following comparisons were not significantly different: Baseline 2 and Post-Injury 3 and Baseline 1 and Post-Injury 1.

Boxplot

Outliers

Of the 19 scores identified to be outliers, none are considered extreme.

Normality Assumption

From the plot, normality is assumed.

ANOVA Computation

## ANOVA Table (type III tests)
## 
##   Effect  DFn    DFd      F        p p<.05 ges
## 1   test 3.82 967.55 82.591 3.61e-58     * 0.1

Pairwise Comparisons

Result Visualization

Baseline 1 Summary

Baseline 1 Histogram

Post-Injury 1 Summary

Post-injury 1 Histogram

Baseline 2 Summary

Baseline 2 Histogram

Post-Injury 2 Summary

Post-injury 2 Histogram

Baseline 3 Summary

Baseline 3 Histogram

Post-Injury 3 Summary

Post-injury 3 Histogram

Baseline and Post-Injury Summary